Warren Buffett Just Confirmed That Berkshire's Alphabet Stake Was His Idea. Why the Stock Still Looks Like a Buy

Buffett’s endorsement is useful mainly as a signal that the market may be underpricing Alphabet’s ability to turn AI capex into a durable cost moat. If TPUs continue to improve training and inference economics, the real upside is not just more AI revenue but higher incremental margins and a better chance that Alphabet can fund price competition without destroying returns on capital. That supports the stock over a 6-18 month horizon, but it is not a fresh fundamental catalyst by itself, so the immediate move is more sentiment-driven than earnings-driven.

The second-order winner is AVGO if custom-silicon adoption expands, because once hyperscalers commit to specialized chips, the attach opportunity spreads into networking, packaging, and platform services. NVDA is the marginal loser only at the margin: this does not break the GPU demand story, but it increases the odds that some AI workloads migrate from merchant silicon to in-house architectures, which can cap multiple expansion if investors start extrapolating Alphabet’s cost structure across the cloud stack.

The contrarian risk is that the market overreads a reputation event and underreads reinvestment risk. If Alphabet’s lower compute cost is used to accelerate product pricing, search monetization, or cloud share gains, reported revenue can look fine while near-term margins remain capped; the falsifier is a cloud operating margin miss or an upward capex revision without corresponding revenue acceleration over the next 1-2 quarters. Berkshire’s own stock may see a small halo bid, but this is not a strong enough signal to justify a standalone trade there.